Demo of automation triggered by SevOne Machine learning insights - Automate config scripts without scripting knowledge

One of the key benefits of SevOne Automated Network Observability is the ability to trigger network automation based on SevOne Machine Learning insights and the ability understand what is normal and what is not. 

An example of this would be the ability to automate config scripts without scripting knowledge. Go to the end of this library entry for a video demonstration. 

Problem

• Configuring QoS perfectly at the first chance is almost impossible, as there will be always some tweaking required. 
• In complex networks with thousands of interfaces and QoS configurations, detecting where QoS could be optimized is very difficult task. 
• In this use-case there is one class that is having lots of packet drops. 
• On the SevOne NPM dashboards there is a clear correlation of the QoS bytes maxing out and packet loss.

Solution

• Use SevOne NPM to analyze all captured data and understand normal behaviour and detect QoS misconfigurations. 
• Using automated workflows we will be able to reconfigure all the misconfigurations automatically in a consistent way. 
• As seen in the image, when the traffic is lower, the amount of packet loss reduces. Indicating that more bandwidth will reduce packet loss.
• In order to detect this type of issue automatically, a policy is created to trigger an alert when there are consistent QoS drops on important QoS classes. 
• A workflow is triggered using a webhook on the alert with the device details provided as input.
• The workflow goes back to the device and QoS class, gets the normal (CurrentBW) QoS traffic out and increases it by 10% (NewBW).
• All the packets that were previously dropped now is sent without the intervention of a person.

Benefit

• Extend automation across your network using the pre-build building blocks and leveraging pre-built workflow templates or build your own automation.
• Reduce downtime by proactively adding closed loop automation to you network powered by ML observations.
• Incorporate third party platforms such as ITSM software or simple email communication to keep track of automated changes in your network. 
• Low-code workflows with advanced API abstraction allows for easy adaption to your specific network and faster time to closed loop automation.